the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 19 Feb 2026
70 Years of Glacier Loss on the Nevados de Chillán volcanic complex, Chile
Abstract. Glaciers on the Nevados de Chillán volcanic complex are rapidly retreating, with anticipated consequences for agroforestry, tourism, and regional human and ecological security. Quantifying their mass balance is critical for understanding current meltwater contributions and for anticipating future water availability as these glaciers continue to shrink. Here we estimate the geodetic mass balance of all 28 documented glaciers on the Nevados de Chillán complex. An uncrewed aerial vehicle (UAV) campaign conducted in March 2024 provided updated elevation data for 11 glaciers on the complex, allowing calculation of volume change from 1954–2024 (70 years). For the remaining 17 glaciers, we analyzed airplane and satellite digital elevation models (DEMs) to estimate volume change from 1954–2000 (46 years). Our results show a clear acceleration in glacier mass loss after 2000 for the glaciers surveyed with UAV data. Mean annual specific mass balance of the Cerro Blanco subcomplex accelerated from -0.41 ± 0.33 m w.e. y-1 (1954–2024) to -0.60 ± 0.29 m w.e. y-1 (2000–2024), while that of the Las Termas subcomplex increased from -0.13 ± 0.32 m w.e. y-1 (1954–2024) to -0.36 ± 0.18 m w.e. y-1 (2000–2024). Regional water resource planning should consider how increasing glacier melt rates on the Nevados de Chillán complex will impact the timing and volume of future water availability.
- Preprint
(9638 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 17 May 2026)
- RC1: 'Comment on egusphere-2025-5795', Robert McNabb, 10 Apr 2026 reply
Model code and software
Glacier-DEM-coregistration-and-MB Millie Spencer and Emma Tyrrell https://doi.org/10.5281/zenodo.17664874
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 207 | 230 | 17 | 454 | 22 | 26 |
- HTML: 207
- PDF: 230
- XML: 17
- Total: 454
- BibTeX: 22
- EndNote: 26
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
In this manuscript, the authors have presented an analysis of 70 years of glacier mass loss in the Nevados de Chillán volcanic complex in south-central Chile. The authors use digitized historic maps and modern datasets to calculate glacier mass balance over two time periods, and compare these results to available meteorological and hydrological data. Overall, I think this is a really interesting study that makes good use of older datsets to help extend the observational record of glacier mass changes further back in time, at least for the provided study area. Overall, I think the study is generally well-written, though I have some general/"major" comments about some aspects of the methods. I also have some smaller/more specific comments, denoted by the line numbers below.
general comments
-----------------
1. I think some care needs to be taken with the vertical datum used by each of your two DEMs, especially the 1954 topographic map DEM. The SRTM that you are using is provided as the height above the EGM96 geoid, while the 1954 map is most likely the same as what is reported by Farías-Barahona et al. (2020), orthometric height (m a.s.l.) measured from tide gauges along the coast. I think there's a good chance of large differences between the vertical datum of the topographic maps and the EGM96 geoid, which would likely show up as differences (errors) on the off-glacier areas of your DEMs, even after co-registration. It might be comparatively minor, but I think it's at least worth mentioning in the data description.
That said, I think you might also want to mask your "source" DEM for the co-registration - it looks as though there are areas outside of the DEM that are still being included for the co-registration, which may also have an impact here.
2. I don't think the approach taken in §3.5, to convert the off-glacier NMAD value to m w.e. and use this as the uncertainty for the mass balance is correct. Because the mass balance is computed as an aggregation of the elevation change, the uncertainty of the volume change needs to be calculated by propagating the uncertainty of a single elevation change pixel through the spatial aggregation - see, for example, Hugonnet et al. 2022 or Rolstad et al. 2009. Because you are using python for some aspects of this (thank you for including your github repository link, and the zenodo link!), you might want to look at the xDEM python package (https://xdem.readthedocs.io/en/stable/; xDEM Contributors, 2024), which includes tools for estimating the DEM uncertainty and propagating that uncertainty to volume changes.
3. Were you able to digitize the glacier outlines from the topographic map? I know that not every topographic map will include glaciers as a distinct feature, but given that other studies such as Farías-Barahona et al. (2020) were able to include the glacier outlines digitized from similar maps, I figure it's at least worth asking about. If glaciers were not identified on the map used, this should be mentioned in the appropriate section as an explanation for why it is not included. If it is possible to include the 1954 outlines, I think it's worth re-doing at least part of the analysis by comparing the mass changes computed with these outlines.
minor/specific comments
------------------------
l. 101-102: I would not characterize these as "images" - the IGM DEM you describe is based on topographic maps which were generated from many images, but what you describe is not using those images. Similarly, the SRTM DEM is not an "image", but rather a DEM.
l. 145: why would manual flying impact the accuracy of the DEM?
Table 1: as at l. 101 - I would describe this as a topographic map, rather than an aerial photo
l. 162: should this be two files, if both points and lines were digitized?
l. 163: what is the density/spacing of the vertices for the polylines? Is this consistent throughout the area digitized?
l. 165: why IDW, rather than a method like Topo to Raster (https://pro.arcgis.com/en/pro-app/3.5/tool-reference/spatial-analyst/topo-to-raster.htm)?
l. 171-179: I don't think this is a correct characterization of the approach taken by both Dussaillant et al. (2019) or Hugonnet et al. (2021). Both of those approaches used many ASTER DEMs, but not to create a single composite DEM as you seem to be suggesting ("ASTER represents a composite elevation from ~2000-2004..."). This is perhaps a correct characterization of the ASTER GDEM (e.g., Abrams et al., 2015), but this is not the same as what was done for those studies. There are issues with ASTER DEMs over steep (north-facing) slopes or low-contrast areas such as snow and ice, but unknown/unquantified signal penetration in the SRTM DEM also poses an issue for estimating glacier mass balance (see, e.g., Berthier et al. 2018). I am also not sure that characterizing the SRTM C-band DEM as "comparing images taken seconds apart" is correct, given that it is a mosaic of acquisitions over an 11-day period (Rabus et al., 2003).
l. 186: neither Shean et al. 2023 nor Berthier et al. 2024 are included in your references.
l. 200: Use MB here, rather than Mass - Balance
l. 201: include a citation for the chosen snow/ice density value
l. 258: Glaciar Nevado
Fig. 3: Showing this as a stretched raster, rather than classified, would make it easier to see the local variation in elevation difference. I would also show the elevation difference of off-glacier areas, though this could be done with some transparency to allow more focus on the on-glacier differences.
l. 290: what is the spatial scale over which the differences are autocorrelated?
Fig. 4: change the labels in the legend to be a line, rather than the outline of a rectangle. Since you are using matplotlib for the figures (thank you for including the github repository!), you can do this by using matplotlib.lines.Line2D rather than matplotlib.patches.Patch for the legend handle - see an example here: https://matplotlib.org/stable/gallery/text_labels_and_annotations/custom_legends.html
l. 300-304: how does the % area change compare to the area-averaged elevation difference for glaciers in different sub-complexes? Looking at this might help explain why the sub-complexes are more sensitive to the choice of outline. Did you also compare the area change to the area-averaged elevation/mass change for each glacier?
l. 360: is this meant to be January and February? In Fig. 6-9, it seems as though the trends are only shown if they are significant, and there is no trend shown in Fig. 7a-c. Additionally, include the p-value for the "significantly" here.
Figs. 6-9: show vertical changepoints on all plots where they are reported and include the year in the legend. Also show the slope pre- and post-change, at least for those plots where you are reporting a change.
l. 400: why would the single year show a larger magnitude than multiple years?
l. 407-420: while comparing to other regional studies is valuable, I think you should also compare to other studies that have reported results for these same glaciers. Hugonnet et al. (2021) provide per-glacier mass balance estimates for the glaciers included in this study for 2000-2019, and those data (as both per-glacier time series and elevation change rasters) can be found here: https://doi.org/10.6096/13. Additionally, WGMS have annual mass change estimates for all RGI glaciers here: https://doi.org/10.5904/wgms-amce-2026-02-10. I think it would be worthwhile to include these estimates in your comparison here, given that they cover the same glaciers, rather than reported estimates for broader regions.
l. 418: you could also discuss the value of longer-term local studies like this one - these are comparatively rare worldwide, but especially in the Andes.
l. 433-434: why the disagreement?
l. 448: can you report the surface area change here?
l. 518: I have to think this would be somewhat negligible compared to the uncertainties in the SRTM, which are typically reported as being at least 5 m or so.
References
----------
Abrams, M., Tsu, H., Hulley, G., Iwao, K., Pieri, D., Cudahy, T., and Kargel, J.: The Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) after fifteen years: Review of global products, International Journal of Applied Earth Observation and Geoinformation, 38, 292–301, https://doi.org/10.1016/j.jag.2015.01.013, 2015.
Berthier, E., Larsen, C. F., Durkin, W. J., Willis, M. J., and Pritchard, M. E.: Brief communication: Unabated wastage of the Juneau and Stikine icefields (southeast Alaska) in the early 21st century, The Cryosphere, 1523–1530, https://doi.org/10.5194/tc-2017-272, 2018.
Farías-Barahona, D., Ayala, Á., Bravo, C., Vivero, S., Seehaus, T., Vijay, S., Schaefer, M., Buglio, F., Casassa, G., and Braun, M.: 60 Years of Glacier Elevation and Mass Changes in the Maipo River Basin, Central Andes of Chile, Remote Sensing, 12, 1658, https://doi.org/10.3390/rs12101658, 2020.
Hugonnet, R., Brun, F., Berthier, E., Dehecq, A., Mannerfelt, E. S., Eckert, N., and Farinotti, D.: Uncertainty Analysis of Digital Elevation Models by Spatial Inference From Stable Terrain, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 15, 6456–6472, https://doi.org/10.1109/JSTARS.2022.3188922, 2022.
Hugonnet, R., Brun, F., Berthier, E., Dehecq, A., Mannerfelt, E. S., Eckert, N., and Farinotti, D.: Uncertainty Analysis of Digital Elevation Models by Spatial Inference From Stable Terrain, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 15, 6456–6472, https://doi.org/10.1109/JSTARS.2022.3188922, 2022.
Rabus, B., Eineder, M., Roth, A., and Bamler, R.: The shuttle radar topography mission - A new class of digital elevation models acquired by spaceborne radar, ISPRS Journal of Photogrammetry and Remote Sensing, 57, 241–262, https://doi.org/10.1016/S0924-2716(02)00124-7, 2003.
Rolstad, C., Haug, T., and Denby, B.: Spatially integrated geodetic glacier mass balance and its uncertainty based on geostatistical analysis: Application to the western Svartisen ice cap, Norway, Journal of Glaciology, 55, 666–680, https://doi.org/10.3189/002214309789470950, 2009.
xDEM contributors: xDEM (v0.1.0). Zenodo. https://doi.org/10.5281/zenodo.11492983, 2024.